As has been mentioned, components used by the builders has a profound effect on the sound coming from the speaker.
The SE triode amplifier I’m currently using came in kit form allowing the luxury of not just tube rolling, but component rolling.
Unfortunately, the amplifier does not have a vacuum tube rectifier. The kit came with Schottky rectifiers. I’d done some unrelated work with 3rd generation silicon carbide rectifiers. I know they have been demonized in some circles but perhaps those were earlier versions (1st generation or 2nd generation-"turn off" characteristics and the resultant current wave have improved significantly). I used a version made by ON Semiconductor which I had used in another project. Nice subtle detail emerged making the switch worthwhile. There wasn’t a lot of space to fit 10 rectifiers in TO-220 packages (the Schottky’s were tiny DO-27 packages) plus carefully bending the leads to fit in place of the tiny rectifiers.
The kit came with Takman carbon film resistors. Not a bad choice. I mean what’s the option, Ohmite phenolic body carbon resistors? (if I were looking for a ’50s or ’60s sound)
It was recommended that Mundorf Silver-Gold-Oil capacitors be used to couple stages (not included in the kit). I started off with them. Then curiosity got the best of me and I swapped the Mundorf’s with Jupiter Copper Foil/Paper Wax capacitors. At over $80 each (4 required), it’s little wonder not a lot of amplifier designers/builders specify them. That capacitor change was a watershed moment. After about 25 hours of burn-in, the audio suddenly went from a rather sterile 2D and flat sound (not necessarily bad depending on the rest of the system) to a very nice detailed presentation. I might have considered Miflex KPCU capacitors (similar construction, similar price) except they physically wouldn’t fit in the amplifier (Jupiters just barely fit).
One other recommendation were the cathode bypass capacitors. Finding a capacitor with very low ESR is desirable. Within the past few years, several Japanese capacitor manufacturers began making organic polymer capacitors that have ESR values in the 60-100 milliohm range. That switch added a bit more detail particularly in the mid-range.
The final roll were the output matching transformers. Lundahl makes very well respected transformers for many applications. They had a transformer that matched the 300B plate characteristics. I swapped the output transformers with Lundahls. This was the second watershed moment. The imaging suddenly came alive, the instruments could be placed in the room, the vocals gained a lot of clarity.
I won’t go into the tube rolling exercise but I will say that Telefunken 12AX7s and 12AU7s sound nice. Ultimately the best sound to my ears were RCA 12AX7As (early ’60s manufacture from Cincinnati, OH plant) and RCA 12BH7As (again, mid ’60s from the Cincinnati, OH plant) in place of 12AU7s. If your preamp or amplifier filament supply can stand the extra filament current draw of the 12BH7, the 30% longer plate/grid/cathode adds nicely to the overall sound stage. And I’ll probably get skewered for saying this, but after listening to 5 different manufactures of 300Bs, to my ears, you can’t beat the original Western Electric (1998 date code) version. The runner up was the Takatsuki 300B (worth a visit if you’re in the Kyoto, Japan area-call for an appointment). I’m still waiting on Charlie Whitener to begin releasing the newly manufactured 300B from Rossville (GA) Works (using much of the original tooling/machinery from the WE Kansas City Works).
Now, it must be pointed out that I have invested as much (maybe more) in component upgrade rolling as the amplifier kit originally cost. Not for the faint of heart, definitely. But the rewards, at least for my listening, has made it all worth the effort and, especially, the money.